Biochimica et Biophysica Acta 1569 (2002) 1^9

www.bba-direct.com

Review

Ascorbic acid: much more than just an antioxidant

Oreste Arrigoni *, Mario C. De Tullio

Dipartimento di Biologia e Patologia Vegetale, Universita di Bari, via E. Orabona 4, 70125 Bari, Italy
Received 26 July 2001; received in revised form 25 October 2001; accepted 31 October 2001

Abstract
Vitamin C (ascorbic acid (AA)) is very popular for its antioxidant properties. Consequently, many other important aspects of this
multifaceted molecule are often underestimated or even ignored. In the present paper, we have tried to bring to the foreground some of these
aspects, including the peculiarities of the AA biosynthetic pathway in different organisms, the remarkable function of AA as a co-substrate
of many important dioxygenases, the role of AA-regenerating enzymes and the known pathways of AA catabolism, as well as the intriguing
function of AA in gene expression. 2002 Elsevier Science B.V. All rights reserved.
Keywords : Ascorbic acid ; Dehydroascorbic acid ; Oxoglutarate-dependent dioxygenase

1. Introduction
Since its discovery in the late 1920s [1], probably no
other chemical has ever been as celebrated as ascorbic
acid (AA). The benecial eect of vitamin C is almost
universally recognised. The reason for this unprecedented
popularity is probably linked on one hand to common
sense (AA is present in relatively high amounts in fruit
and vegetables, which are known to be healthy), and on
the other hand ^ especially today ^ to expensive advertising campaigns for vitamin C-based products.
Until a few decades ago, the most common answer to
the question `what is the function of vitamin C?' would
have been the same given by Albert Von Szent Gyorgyi in
1937, when he was awarded the Nobel Prize for the discovery of AA: it is the factor able to cure the variety of
clinical symptoms known as scurvy, a syndrome occurring
in humans whose diet is decient in fresh fruit and vegetables. At that time, however, little molecular explanation
for this eect was available. Nowadays, the same question
would receive a dierent answer. Not many would mention scurvy, as this pathological state is no longer very
common. Some specialists in the eld would explain that
the inability of humans to synthesise AA is due to the lack
of L-gulono-lactone oxidase (GulL-ox), the last enzyme in
* Corresponding author. Fax: +39-080-5442158.
E-mail addresses : arrigoni@botanica.uniba.it (O. Arrigoni),
detullio@botanica.uniba.it (M.C. De Tullio).
1
Also corresponding author : Fax: +39-080-5442158

its biosynthetic pathway. A large majority of people asked

(irrespective of their scientic education) would promptly
answer that AA is an antioxidant which eciently scavenges toxic free radicals and other reactive oxygen species
(ROS) formed in cell metabolism. Actually, ROS are associated with several forms of tissue damage and disease
and also with the process of ageing [2]. Aerobic organisms
have evolved intricate and interrelated processes for protection against the eects of free radicals and derived toxic
species, including both enzymatic and non-enzymatic defences. Enzymatic mechanisms include superoxide dismutase (SOD), catalase (CAT), glutathione peroxidase (GSHpx) and, in plants, ascorbate peroxidase (AA-px).
According to the generally assumed model of enzymemediated removal of ROS [3], SOD catalyses the disproportionation of superoxide anion to H2 O2 and oxygen; in
turn H2 O2 is converted by CAT into water and molecular
oxygen. CAT turnover number is very high, but its anity
for H2 O2 is relatively low, and consequently a certain
amount of H2 O2 remains in the cell. This is potentially
troublesome, since H2 O2 can react with superoxide anion
formed in oxidative metabolism and generate the highly
reactive hydroxyl radical. GSH-px and AA-px are capable
of removing low amounts of H2 O2 due to their high anity for H2 O2 . Thus, the co-operativity of SOD, CAT and
peroxidases ensures low levels of superoxide anion and
H2 O2 and therefore limits the risk of hydroxyl radical
formation (Fig. 1). In addition to enzymatic mechanisms,
some compounds, known as antioxidants, are present in
the cell that, on entering into redox reactions, contribute

0304-4165 / 02 / $ ^ see front matter 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 3 0 4 - 4 1 6 5 ( 0 1 ) 0 0 2 3 5 - 5

BBAGEN 25270 23-1-02

Cyaan Magenta Geel Zwart

O. Arrigoni, M.C. De Tullio / Biochimica et Biophysica Acta 1569 (2002) 1^9

The aim of the present article is to summarise our

present knowledge on the AA system in both animals
and plants, in order to demonstrate that the pleiotropic
action of vitamin C is, in all organisms, largely due to its
function as a co-substrate of dioxygenases, a versatile family of enzymes. Moreover, we have also tried to suggest
new directions for research, in order to give a three-dimensional portrait of this very well known, but still poorly
understood compound.
2. Biosynthesis and evolution
Fig. 1. The role of AA in the detoxication of ROS. Blue dotted lines
indicate non-enzymatic reactions.

an electron to neutralise free radicals to non-reactive species. Many chemicals could serve this purpose because the
high reactivity of free radicals results in extracting an electron from almost any available molecule. However, an
ecient biological antioxidant is supposed to do more
than simply react with free radicals: (a) it must be present
in an adequate amount in the cell, (b) it must react with a
variety of free radicals, and (c) it must be suitable for
regeneration [4]. This combination of properties is typical
of AA ; therefore this compound is considered the perfect
antioxidant for the cells of nearly all aerobic organisms.
However, it should be considered that under several circumstances ROS and free radicals in general also play a
pivotal role in signal transduction and the mechanism of
enzyme action [5,6], thus indicating that the function of
antioxidants is the ne regulation, rather than total extinction of free radicals. Moreover, in some cell compartments
(i.e. plant cell walls) AA regeneration is impaired due to
the absence of NAD(P)H and GSH. Eventually, if AA can
be considered an ecient antioxidant and consequently an
ecient free radical scavenger, it cannot be ignored that
under some circumstances (such as the presence of metal
ions and adequate pH conditions) AA has a pro-oxidant
and even mutagenic eect [7^9].
All the above-mentioned observations lead us to conclude that deciphering the details of the antioxidant action
which is attributed to vitamin C it is still very dicult for
at least two reasons: (a) we lack reliable analytical methods to accurately measure AA distribution in dierent cell
compartments, and (b) AA antioxidant action takes place
by means of non-enzymatic (and therefore more dicult
to predict) reactions, excepting AA-px. In addition, considering the key role of AA in cell metabolism to be limited to its antioxidant action could be misleading. In fact,
this approach does not take into consideration many AAdependent metabolic reactions that inuence essential
physiological processes, ranging from cell division to
gene expression and the activation of biological defence
mechanisms [10].

BBAGEN 25270 23-1-02

Biosynthesis of AA was a relatively late event in evolution. Apparently neither AA nor AA-related enzymes are
present in Escherichia coli or other bacteria [11]. The presence of D-araboascorbic acid was detected in Penicillium
[12], whereas Saccharomyces cerevisiae and Candida albicans can synthesise the 5-carbon AA analogue D-erythroascorbic acid (EAA) [13] via D-arabinose and D-arabinono-1,4-lactone [14]. All plants can synthesise AA at high
rates following a four-step pathway (the Smirno^Wheeler
pathway) which proceeds from GDP-D-mannose, via
GDP-L-galactose, L-galactose and L-galactono-1,4-lactone
(GalL), but possible minor biosynthetic pathways have
not been excluded [15]. Almost all animals synthesise
AA via UDP-D-glucuronic acid, glucuronic acid/glucuronolactone, L-gulonic acid/L-gulono-1,4-lactone [16,17],
with the exception of humans, some primates, guinea
pigs, some birds and teleost shes [18]. The nal steps of
the three main pathways of AA synthesis are compared in
Fig. 2.
It has been suggested that AA biosynthesis rst appeared in amphibians, in connection with the passage of
early tetrapods from an aquatic to a terrestrial environment, where oxygen could actually be a threat [18]. However, this hypothesis does not t with more recent ndings
on the presence of AA biosynthetic enzymes in aquatic
organisms, such as archeopterigian shes and lamprey
[19,20]. In support of the hypothesis that AA biosynthesis
evolved side by side with the increase in atmospheric oxygen, it has been observed that animals which have lost the
capability to synthesise AA (guinea pig, bat, monkey and
man) show higher SOD activities in both liver and kidneys
[21]. Following this line of reasoning, however, cyanobacteria, being the rst organisms to produce O2 , should also
have been the rst ones to sense it and to nd a way to
avoid ROS accumulation; in this case, AA would have
been a useful tool. Unfortunately, scarce and contradictory information is available concerning AA biosynthesis
and metabolism in cyanobacteria [22].
It is now clear that man and other animals cannot synthesise AA because the gene encoding L-gulono lactone
oxidase (GulL-ox), the enzyme catalysing the last step in
AA biosynthesis, is not functional anymore [23]. Rat
GulL-ox expression in guinea pig cells yielded AA [24].

Cyaan Magenta Geel Zwart

O. Arrigoni, M.C. De Tullio / Biochimica et Biophysica Acta 1569 (2002) 1^9

Fig. 2. A comparison of EAA and AA biosynthesis in dierent organisms. Only the nal two steps occurring in yeasts, animals and plants
are reported. ARA : D-arabinose; GUA : D-gulonic acid ; GAL: L-galactose; AL: D-arabinono lactone; GulL: L-gulono lactone ; GalL: L-galactono lactone. EAA: D-erythroascorbic acid ; AA: L-ascorbic acid. GulL
conversion to AA occurs in the ER, whereas GalL conversion occurs in
mitochondria (Mit).

This conrms earlier reports showing that previous steps

in the biosynthetic pathway are still operating in organisms which do not synthesise AA [17], although it is not
clear whether the enzymes catalysing such activities are
involved in other metabolic pathways. Surprisingly, this
point has almost completely been neglected by scientists.
It has been suggested that the loss of function of the
GulL-ox gene, and the consequent impaired synthesis of
the antioxidant AA, increased the mutation rate in the
genome of our ancestors, inducing faster evolution of
Anthropoidea and Homo sapiens [25,26]. On the other
hand, the opposite could also be true: since GulL-ox activity produces hydrogen peroxide, it cannot be excluded
that the loss of GulL-ox activity could have been a favourable trait in terms of evolution [9]. The fact that all other
enzymes in the animal biosynthetic pathway are still
present in guinea pig and man could be in line with this
hypothesis. It should also be considered that even in AAsynthesising organisms AA is apparently not produced
during early life stages, as in the case of newborn rats
[27] and in developing seeds of plants [28].
The location of AA biosynthetic enzymes is another
interesting issue. In primitive shes, amphibians and reptiles, AA synthesis takes place in kidneys, whereas the liver
is the site of synthesis in mammals [20,21]. A more complex situation is observed in birds, where all three condi-

BBAGEN 25270 23-1-02

tions have been observed (kidney, liver, or no synthesis)

[29]. In plants, virtually all cells synthesise AA, but the
synthesis is much higher in juvenile (meristematic) cells
[30].
Subcellular location of AA biosynthetic enzymes also
shows peculiar features in dierent organisms. In plants,
the rst steps of AA synthesis are likely to be cytosolic,
whereas GalL dehydrogenase, catalysing GalL conversion
into AA, is located at the inner mitochondrial membrane
[31]. Similarly, AA biosynthesis in animals is carried out
by soluble enzymes, apart from the last step catalysed by
GulL-ox, which is located in the endoplasmic reticulum
(ER) [16,21].
It has been suggested that the transfer of the AA biosynthetic machinery in animals from the kidney of coldblooded reptiles to the much larger and more active liver
of mammals reected the increased need for AA due to
increasing atmospheric oxygen [21]. Whatever the reason
for this change, it is surprising that a compound so very
useful in high concentrations in all organs and cell compartments (the more, the better !) is only synthesised in a
single organ. The comparison with hormones, synthesised
in a specic site and then transferred to their dierent
targets, is clearly not consistent, as hormones are eective
at very low concentrations and, in general, hormone targets are represented by a few cells responding to the hormone stimulus. On the other hand, the reason for localised
AA synthesis could reside in the necessity to keep it under
tight control, suggesting that generalised and unleashed
AA synthesis could be as dangerous as AA deciency.
The direct consequence of limiting to a single organ the
synthesis of an essential compound, required for the metabolic activity of all cells, has been the development of
short- and long-range ecient AA transport. The presence
in AA-synthesising organisms of a specic delivery system
capable to chaperone AA and avoid oxidative processes
during its transport has been hypothesised [10], but as yet
there is no direct evidence of such a mechanism.
The evolutionary reason for the presence in mitochondria of the nal step of the AA biosynthetic pathway in
plants is not clear. It is known that AA synthesis is linked
with the mitochondrial electron transport chain and that
GalL dehydrogenase seems to utilise cytochrome c as a
specic electron acceptor [31], but this point requires further investigation. As discussed below, the location in the
ER of the last step of AA biosynthesis in animals is a
remarkable example of biological organisation.
More information is available concerning exogenous vitamin C uptake, which is very important in non-synthesising organisms. In animals, both AA and dehydroascorbic
acid (DHA) are actively transported. However, the oxidised vitamin C form DHA is preferentially transported
across plasma membranes by means of the GLUT1 glucose transporter [32].
In plant cells DHA is also actively transported across
the plasma membrane [33]. This mechanism is tightly as-

Cyaan Magenta Geel Zwart

O. Arrigoni, M.C. De Tullio / Biochimica et Biophysica Acta 1569 (2002) 1^9

sociated to DHA reduction to AA and seems to be less

relevant in the framework of the AA system. In plants, all
cells (both photosynthetic and non-photosynthetic) can
synthesise AA de novo from sugar precursors, therefore
the exchange of AA and DHA between cells is less important. It is possible to hypothesise an AA long-range transport mechanism in plants from young leaves and reproductive organs. This is consistent with the observation that
the Vicia faba embryo has a high AA content despite not
being able to synthesise AA until 30 days after anthesis, as
GalL dehydrogenase is not functional at this stage [28].
Developing seeds acquire the capability to oxidise GalL to
AA just 10 days before the onset of the desiccation period.
A few days before entering the resting state, the seed is
endowed with the entire AA biosynthetic pathway, so that
the seed can promptly synthesise the amount of AA
needed to re-start metabolic activity at the onset of germination. From this stage on, the AA biosynthetic machinery is always functional, and appears to be regulated by a
feedback mechanism (Fig. 3). Freshly harvested potato
tubers show high endogenous AA content and are not
capable of synthesising AA when fed with glucose. AA
content gradually decreases during storage, along with
an increase in the capability to synthesise AA from glucose
[34]. Similar data have recently been obtained by studies
with 14 C-labelled glucose in pea seedlings pre-loaded with
AA [35]. Furthermore, GalL administration strongly induces AA synthesis irrespective of endogenous AA content, clearly indicating that the feedback mechanism operates upstream of GalL dehydrogenase (Fig. 3).
3. AA-dependent dioxygenases (AADs)
Dioxygenases catalyse the incorporation of O2 into an
organic substrate. This class of proteins includes a wide
variety of enzymes, which dier in substrate requirement
and mechanism of action. A subgroup of dioxygenases
require oxoglutarate for their catalytic activity, and are
therefore usually referred to as `2-oxoglutarate-dependent
dioxygenases' [36]. In turn, some of these enzymes specifically require AA, and at least in one case (the plant enzyme 1-aminocyclopropane-1-carboxylate oxidase) AA is
specically required as a co-substrate, whereas 2-oxoglutarate is not [37]. Hence, from this point on, we will refer
to these enzymes as AADs. In addition, extending to all
AADs the nomenclature previously used by Kivirikko [38]
to describe the mechanism of prolyl-hydroxylase (see below), we suggest the use of the term co-substrate for AA,
rather than the commonly used cofactor.
The identication of AA as a co-substrate of AADs is a
long story, taking its origins from the circumstantiated
clinical observations of James Lind, who in 1747 observed
that an unknown compound present in fruit and vegetables was able to cure scurvy [39]. It took two more centuries to identify AA as the antiscorbutic factor, and some

BBAGEN 25270 23-1-02

Fig. 3. The feedback mechanism of AA biosynthesis in plants. Freshly

harvested potato tubers were stored and their endogenous AA content
(red line) measured at dierent time points. Biosynthetic capability was
measured in potato tuber slices fed for 24 h with 5 mM D-glucose (blue
line). Administration of 5 mM GalL to slices for 24 h always markedly
increased AA content (red bars). Modied from [34].

more decades to understand its biochemical action. By the

early 1960s, AA involvement in the processes of collagen
synthesis, tyrosine oxidation, corticosteroid synthesis and
aromatic hydroxylation had been more or less ascertained
([40] and references therein), but the biochemical basis of
AA functions would only be fully understood some years
later, starting with the discovery of the post-translational
modication of peptidyl-prolyl residues by prolyl-4-hydroxylase (P4H) [41]. Data on the specicity of AA requirement for P4H activity were questioned by Udenfriend
and co-workers on the basis of in vitro experiments, in
which other agents were capable of taking over AA function in collagen hydroxylation [42]. Further experiments
have demonstrated that inhibition of P4H activity in
AA-synthesising organisms induces an increase in AA content in vivo [43,44], thus clearly indicating that P4H activity actually consumes AA.
Hydroxylation at carbon 4 of proline residues incorporated in polypeptide chains is a post-translational modication which is extremely important for protein folding
and secretion. In vertebrates, P4H is a tetramer made up
of two K- and two L-chains. The hydroxylation reaction is
catalysed by K-subunits, whereas the role of L-subunits is
not yet clear. Notably, it has been ascertained that L-subunits and protein disulde isomerase (PDI) are the product of the same gene [38], but at present the role of the
PDI protein as a subunit of P4H is not clear. The association between K- and L-subunits is much less constant in
the plant enzyme [45]. P4H resides in the ER and recognises specic proline-containing sequences [38].
Hydroxyproline residues are typically present in proteins of the extracellular matrix : collagen in animals,
and an array of dierent plant proteins (generically known

Cyaan Magenta Geel Zwart

O. Arrigoni, M.C. De Tullio / Biochimica et Biophysica Acta 1569 (2002) 1^9

as hydroxyproline rich glycoproteins, HRGPs) ranging

from structural elements to proteins involved in developmental processes [46]. Interestingly, in the absence of ecient hydroxylation of peptidyl-prolyl residues, all these
proteins are non-functional and are rapidly degraded in
the ER, whereas correctly hydroxylated proteins are exported and localise in the extracellular matrix [47]. In animals, collagen underhydroxylation is not only the main
cause of scurvy, but also of other clinical disorders such
as Ehlers^Danlos syndrome [48]. In plants, the inhibition
of P4H activity induces dramatic alterations to cell wall
regeneration and cell division in tobacco protoplasts [49]
and markedly aects cell morphology and development of
onion roots [44].
Collagen maturation is a complex multi-step process, in
which, after protein synthesis and hydroxylation of peptidyl-prolyl residues, stable trimers are formed by a mechanism involving thiol/disulde exchanges [50]. It is noteworthy that, in animals, AA biosynthesis, P4H activity
and collagen maturation occur in the same cell compartment (the ER). A scheme of all these activities related to
the AA system is reported in Fig. 4. As mentioned above,
this intricate network of functions and activities is a fascinating example of biological organisation. We hypothesise the co-operative action of the L-subunits of P4H and
the ascorbate system to ensure correct folding and the
formation of the triple helix of collagen. The hydroxylation process catalysed by P4H K-subunits, utilising AA as
an electron donor, generates ascorbate free radical (AFR,
see below). This radical is in part recycled to AA by AFR
reductase, which is present in the ER, but part of AFR is
thought to escape reduction and spontaneously disproportionate yielding DHA. Such DHA could be the physiolog-

ical electron acceptor necessary to form the disulde

bonds necessary for protein folding. Moreover, it has recently been reported that both PDI and the heat shock
protein HSP47 bind to underhydroxylated procollagen
synthesised in the absence of AA and display co-operative
but distinct chaperone functions during procollagen biosynthesis [51].
Beside P4H, other AADs are aected by AA deciency.
Carnitine synthesis requires the sequential action of two
AADs: O-N-trimethyl-L-lysine hydroxylase and Q-butyrobetaine hydroxylase [52]. In addition, AA administration
induces carnitine synthesis in guinea pigs [53]. Weakness,
which is the rst detectable symptom of scurvy, has been
correlated with incorrect functioning of these enzymes. We
have already mentioned that AA involvement in `tyrosine
oxidation' has been known for ve decades [54]. This reaction is catalysed by tyrosine hydroxylase and represents
the rate-limiting step in the biosynthesis of the catecholamines dopamine, norepinephrine, and epinephrine [55]. 4Hydroxyphenylpyruvate dioxygenase catalyses the formation of homogenisate from 4-hydroxyphenylpyruvate and
molecular oxygen [56]. Aspartyl (asparaginyl) L-hydroxylase activity is responsible for the post-translational modication of Asp residues present in EGF-like domains [57].
AADs operate in plants for the synthesis of avonoid
and important hormones: gibberellic acid, ethylene and
abscisic acid (ABA) [58]. The pathway of ABA biosynthesis [59] has striking similarities to the synthesis of carotenoids involved in visual processes [60]. Deoxyhypusine
hydroxylase, an enzyme catalysing the last step of the
post-translational modication of a specic lysine residue
forming the unusual amino acid hypusine, is present in the
eukaryotic initiation factor 5A [61]. Interestingly, inhibi-

Fig. 4. The AA system in the ER. See text for details.

BBAGEN 25270 23-1-02

Cyaan Magenta Geel Zwart

O. Arrigoni, M.C. De Tullio / Biochimica et Biophysica Acta 1569 (2002) 1^9

tion of deoxyhypusine hydroxylase activity blocks the G1^

S transition of the cell cycle [62].
The list of AADs identied so far is quite long, and it is
conceivable that some more will be identied in the near
future. An exhaustive listing of known AADs is far beyond the scope of this article. However, we would like to
stress that so many dierent enzymatic functions all requiring AA may well be the basis to understand the pleiotropic action observed for this compound.
4. AFR and DHA reduction to AA: simple regeneration or
more complex role?
Dioxygenase activity and all other AA-requiring reactions (either enzymatic and non-enzymatic) generate AFR
(also referred to as monodehydroascorbate radical). Several reports indicate a specic biological role for AFR,
which is apparently involved in growth processes in both
plant and animal cells [63,64]. It has been suggested that
such AFR action could be mediated by a trans-plasma
membrane electron transport system [65]. It is unclear
whether other enzymatic AFR-requiring reactions actually
occur, but this eld has not been fully explored. On the
other hand, it is well-established that AFR spontaneously
disproportionates yielding DHA (and AA). At the present
time, there is no evidence of reactions catalysing direct
oxidation of AA to DHA, therefore intracellular DHA is
apparently exclusively due to AFR disproportionation.
Besides disproportionation, there is another possible
fate for AFR: it can be reduced back to AA by NADHdependent AFR reductase (EC 1.8.5.3). This enzyme was
identied in dierent organisms in the 1950s [66^70]. As
for the location of AA biosynthetic enzymes, the subcellular distribution of AFR reductase in dierent organisms
is an intriguing topic. In animals, AFR reductase activity
is known to be located both in the ER, mitochondria [71]
and plasma membrane, the latter depending on cytosolic
NAD(P)H in an electron transport system mediated by
CoQ inserted in the bilayer [72,73]. In plants, AFR reductase is located in the cytosol, chloroplasts and mitochondria [70,74,75]. As mentioned above, the AFR generated
by AA utilisation that escapes AFR reductase disproportionates generating both AA and DHA. The latter can be
reduced back to AA either non-enzymatically or by enzyme-mediated mechanisms. In both cases, glutathione
(GSH) has a prominent role as an electron donor [27].
The concomitant presence in chloroplasts of AFR reductase, DHA reductase and glutathione (GSH) reductase,
suggested for plants a fascinating AA regeneration pathway known as the AA^GSH cycle [75], in which AFR and
DHA generated as a consequence of AA utilisation (i.e.
oxidation) are continuously recycled to AA.
However eciently AA regeneration from AFR and
DHA may occur, each cell and obviously any growing
organism requires AA biosynthesis (or dietary intake in

BBAGEN 25270 23-1-02

non-synthesising organisms) because AA and DHA undergo continuous degradation. We believe that AFR reductase plays a key role in limiting DHA formation, rather
than AA generation, as DHA is potentially toxic and a
powerful inhibitor of several enzymes [76^78]. High AFR
reductase activity limits the amount of AFR available for
the disproportionation reaction. This hypothesis is supported by the observation that actively dividing cells,
which use large amounts of AA for their metabolism
and for division processes, also show higher AFR reductase activity, lower DHA content and highly ecient de
novo AA biosynthesis [30].
DHA can also be eciently reduced back to AA. However, the presence of a specic DHA regeneration system
is another controversial point. In animal tissues, the presence of specic enzymatic mechanisms has been repeatedly
questioned [79,80]. GSH-dependent DHA reductase activity has been detected in many plant species, and enzymes
showing DHA reductase activity have been puried from
dierent sources [81]. Since the observation that glutaredoxin and PDI can also catalyse DHA reduction to AA
[82], the list of animal and plant proteins that share DHA
reductase activity, besides catalysing dierent reactions, is
getting longer and longer, and presently also includes thioredoxin and Kunitz-type trypsin inhibitor [83], thioredoxin reductase [84], 3K-hydroxysteroid dehydrogenase [85],
GSH peroxidase [86], dioscorins [87] and even rat serum
albumin [88]. This list does not include other reports on
enzymes with DHA reductase activity, the sequences of
which are not related to previously described proteins
[89^91]. This multiplicity of DHA-reducing enzymes is apparently linked to the peculiar reactivity of the typical
dicysteinyl motif (C-X-X-C), which is present in virtually
all the above-mentioned proteins sharing DHA reductase
activity.
The physiological role of DHA reductase is still a matter for debate. In animals lacking the AA biosynthetic
machinery, AFR and DHA reduction could in principle
contribute to keeping the size of the AA pool constant.
However, the overall activity of these two enzymes is not
much dierent in AA-synthesising and non-synthesising
organisms, therefore it is conceivable to conclude that
the bulk of AA in all organisms results from de novo
biosynthesis or dietary intake, and not from AFR and
DHA recycling. This has been observed, for instance,
when germinating maize embryos were detached from
their storage tissues, thus reducing the availability of sugar
AA precursors [92]; in spite of high AFR and DHA reductase activities, AA content was progressively slowed
down. According to computer simulations, there would
be no need for AA recycling from DHA in chloroplasts
in order to achieve the AA concentrations measured in
this organelle [93]. Although DHA reduction is unlikely
to signicantly contribute to AA content, it could nevertheless have a role in DHA removal. We have recently
observed that DHA administration to lupin and onion

Cyaan Magenta Geel Zwart

O. Arrigoni, M.C. De Tullio / Biochimica et Biophysica Acta 1569 (2002) 1^9

roots induces a remarkable increase in intracellular AA

content, in parallel with a transient decrease in the size
of GSH pool and a general oxidation of thiol-containing
proteins. However, AA formed by means of DHA reduction (either enzymatic or non-enzymatic) is apparently not
equivalent to AA formed via biosynthesis [94]. This is
consistent with the observations that DHA administration
inhibits plant growth and only partially reverses scurvy
symptoms in guinea pigs [95], although DHA is massively
reduced to AA. Therefore, we believe that considering the
AA pool as the sum of AA plus DHA content (as sometimes reported) is incorrect, since many available data indicate that AA and DHA are chemical species with dierent reactivity and metabolic functions, rather than just
two sides of the same coin.
In animal cells, a signicant contribution to DHA reduction by protein thiols has been observed in the ER [96],
further substantiating the hypothesis of a role for DHA in
the regulation of thiol^disulde exchanges and in protein
folding (Fig. 4). The peculiar reactivity of DHA with protein SH groups, beyond its specic cellular context in the
ER, could explain its toxicity and inhibiting eect on plant
growth [94]. This could also give a convincing explanation
of DHA involvement in diabetes [97], possibly linked to
DHA accumulation due to impaired reduction [98,99].
5. AA catabolism and the DHA mystery
Almost all biological research on the AA system in
plants and animals has been directed to the investigation
of its three most representative chemical species, namely
AA, AFR and DHA. Not much attention has been given
to diketogulonic acid (DKG) and to downstream steps of
the degradation pathway. In the rst step of DHA degradation, the lactone ring is broken and DKG is formed
[100]. Attempts to identify an enzyme catalysing DHA
delactonisation have not been successful. It is known
that this reaction is promoted by bicarbonate [101].
Although a clear view of the pathway of AA catabolism
has not been obtained yet, it is known that in animal cells
DKG degrades to 5- and 4-carbon species, including lyxonic acid, xylonic acid, lyxose and xylose [102]. In plants,
Loewus and co-workers identied oxalate, threonate and
tartrate as products of AA catabolism [103]. However,
available information on AA catabolism is far from being
conclusive. In particular, the `DHA mystery' largely remains unsolved. As previously mentioned, AA is massively
utilised (dioxygenase activity, enzymatic and non-enzymatic ROS scavenging). All these reactions generate
AFR, and then DHA as a product of disproportionation.
However, we have observed that DHA content is usually
quite low, as compared to AA [30] and usually it is not
accumulated, even when it is exogenously administered
[94]. The amount of AA degradation products observed
is apparently not consistent with the amount of AA con-

BBAGEN 25270 23-1-02

sumed in cell metabolism [104]. Where does DHA end up?

A possible answer to this question could be obtained from
interesting work in the eld of chemistry.
The stability of DHA solutions may vary according to
pH conditions [100]. Interestingly, analysis of DHA solutions at dierent time points has provided an array of
many dierent molecules [105]. Experimental conditions
used for such experiments on AA degradation in non-biological systems are often very far from physiological conditions, and involve high pH values, high temperatures,
and high DHA concentrations. Nevertheless, these studies
deserve our attention, since at least part of the degradation products isolated by using this approach might be
produced by (yet unknown) catalysed reactions in vivo.
Such hypothetical enzyme-mediated reactions could possibly shed new light on AA catabolism in vivo and help in
solving the mystery of the fate of DHA.
Jung and Wells observed three high-performance liquid
chromatography peaks deriving from DHA solutions kept
at pH 7 and 30C for 10 min, and two of the peaks were
identied as AA and EAA [106]. As previously mentioned,
EAA is a 5-carbon analogue of AA synthesised in yeast
cells [13]. There is little information about a possible biological role of EAA. However, this compound presents the
same enediol group typical of AA and it has been reported
to have AA-like activity in the larval growth of tobacco
hornworm (Manduca sexta) [107] and to protect yeast cells
from hydrogen peroxide [13]. In addition, EAA may function as an electron donor yielding erythroascorbate free
radical (EAFR), and this in turn forms dehydroerythroascorbate. EAFR reductase activity has recently been described in S. cerevisiae [108]. At present, there is no indication that DHA degradation could actually yield EAA in
biological systems, although this hypothesis deserves further investigation.
6. AA-induced mRNA transcription
An increasing number of studies have investigated the
involvement of AA in the regulation of transcription and/
or stabilisation of specic mRNAs. Apparently specic
ascorbate^RNA complexation was observed through
both G^C and A^U base pairs, without changing RNA
secondary structure [109]. This could be related to the
observed selective AA-dependent stabilisation of collagen
transcript and destabilisation of elastin transcript in
smooth muscle cells and skin broblasts [110]. Alternatively, transcription of collagen genes could be induced
by mature (i.e. properly hydroxylated and folded) collagen, therefore AA could have an indirect eect on transcription. As a third hypothesis, a redox-controlled mechanism has been invoked. The transcription of the 72-kDa
type IV collagenase (matrix metalloproteinase-2) is downregulated by AA in cultured human amnion-derived cells
[111]. Tyrosine hydroxylase transcription is enhanced by

Cyaan Magenta Geel Zwart

O. Arrigoni, M.C. De Tullio / Biochimica et Biophysica Acta 1569 (2002) 1^9

AA treatment [112], and the mRNA encoding various

forms of cytochrome P450 in liver microsomes from guinea pigs is induced as well [113]. Other mRNAs whose
transcription appears to be regulated by AA are those
for ubiquitins in guinea pigs [114]. AA alters the expression of collagen integrins in bone culture [115]. Avigliano
and co-workers have recently demonstrated that AA administration induces transcription of the fra-1 gene, which
encodes a transcription factor of the Fos family, and
down-regulates the activator protein-1 (AP-1) target genes
[116]. Moreover, AA has been observed to aect vitamin
D-induced dierentiation of leukemic cells by redox regulation of AP-1, and also aects nuclear factor UB [117]. In
plants, the maize Hrgp gene is induced by AA [118]. It is
noteworthy that at least part of the genes which have their
transcription induced (or stability of the transcript enhanced) by AA encode proteins which contain hydroxylated proline residues or which require AA for their catalytic activity.
7. Conclusions
In an interview published a few years before passing
away, Albert Von Szent-Gyorgyi described AA and its
reactivity as the basis itself of the continuing changes of
life [119]. In the light of our present knowledge, this can be
viewed as a beautiful and stimulating poetical image,
rather than a scientically based theory. As mentioned
in the Section 1, AA has been much studied, yet its potential contribution to many dierent aspects of cell metabolism has been neglected. It is worth mentioning, for
instance, that the regulation of the process of cell proliferation via inhibition of the dioxygenase deoxyhypusine
hydroxylase has recently been viewed as a possible target
for anti-cancer strategies [62,120,121]. It is very easy to
predict that in the near future many more intriguing aspects of the AA system will give rewarding results to researchers interested in their investigation.
Acknowledgements
The authors wish to thank two anonymous referees for
their valuable comments.
References
[1] M.B Davies, J. Austin, D.A. Partridge, Vitamin C: its Chemistry and
Biochemistry, The Royal Society of Chemistry, Cambridge, 1991.
[2] B.N. Ames, M.K. Shigenaga, T.M. Hagen, Proc. Natl. Acad. Sci.
USA 90 (1993) 7915^7922.
[3] I. Fridovich, Ann. N.Y. Acad. Sci. 893 (1999) 13^18.
[4] R.C. Rose, A.M. Bode, FASEB J. 7 (1993) 1135^1142.
[5] R. Schreck, P. Rieber, P.A. Baeuerle, EMBO J. 10 (1991) 2247^2258.
[6] U. Uhlin, H. Eklund, Nature 370 (1994) 533^539.

BBAGEN 25270 23-1-02

[7] I.D. Podmore, H.R. Griths, K.E. Herbert, N. Mistry, P. Mistry,

J. Lunec, Nature 392 (1998) 559^561.
[8] J.H. Song, S.H. Shin, G.M. Ross, Brain Res. 895 (2001) 66^72.
[9] B. Halliwell, Mutat. Res. 475 (2001) 29^35.
[10] M.C. De Tullio, O. Arrigoni, Curr. Topics Phytochem. 2 (1999) 135^
143.
[11] B.L. Lee, W.K. Huh, S.T. Kim, J.S. Lee, S.A. Kang, Appl. Environ.
Microb. 65 (1999) 4685^4687.
[12] T. Takahashi, M. Mitsumoto, H. Kayamori, Nature 188 (1960) 411^
412.
[13] W.K. Huh, B.H. Lee, S.T. Kim, Y.R. Kim, G.E. Rhie, Y.W. Baek,
C.S. Hwang, J.S. Lee, S.O. Kang, Mol. Microbiol. 30 (1998) 895^903.
[14] R.D. Hancock, J.R. Galpin, R. Viola, FEMS Microbiol. Lett. 186
(2000) 245^250.
[15] N. Smirno, P.L. Conklin, F.A. Loewus, Annu. Rev. Plant Physiol.
Plant Mol. Biol. 52 (2001) 437^467.
[16] M. Ul Hassan, A.L. Lehninger, J. Biol. Chem. 223 (1956) 123^138.
[17] N. Smirno, Vitam. Horm. 61 (2001) 241^266.
[18] I.B. Chatterjee, Science 182 (1973) 1271^1272.
[19] K. Dabrowski, Experientia 50 (1994) 745^748.
[20] R. Moreau, K. Dabrowski, Proc. Natl. Acad. Sci. USA 95 (1998)
10279^10282.
[21] A. Nandi, C.K. Mukhopadhyay, M.K. Ghosh, D.J. Chattopadhyay,
I.B. Chatterjee, Free Radic. Biol. Med. 22 (1997) 1047^1054.
[22] C. Obinger, G. Regelsberger, A. Pircher, G. Strasser, G.A. Peschek,
Physiol. Plant. 104 (1998) 693^698.
[23] M. Nishikimi, K. Yagi, Am. J. Clin. Nutr. 54 (1991) 1203S^1208S.
[24] A. Krasnov, M. Reinisalo, T.I. Pitkanen, M. Nishikimi, H. Molsa,
Biochim. Biophys. Acta 1381 (1998) 241^248.
[25] J.J. Challem, Med. Hypotheses 48 (1997) 387^392.
[26] J.J. Challem, E.W. Taylor, Free Radic. Biol. Med. 25 (1998) 130^132.
[27] A. Meister, Biochem. Pharmacol. 44 (1992) 1905^1915.
[28] O. Arrigoni, L. De Gara, F. Tommasi, R. Liso, Plant Physiol. 99
(1992) 235^238.
[29] C.R. Chaudhuri, I.B. Chatterjee, Science 164 (1969) 435^436.
[30] O. Arrigoni, J. Bioenerg. Biomembr. 26 (1994) 407^419.
[31] C. Bartoli, G. Pastori, C.H. Foyer, Plant Physiol. 123 (2000) 335^
343.
[32] J.C. Vera, C.I. Rivas, J. Fischbarg, D.W. Golde, Nature 364 (1993)
79^82.
[33] N. Horemans, H. Asard, R.J. Caubergs, FEBS Lett. 421 (1998) 41^
44.
[34] L. DeGara, F. Tommasi, R. Liso, O. Arrigoni, Boll. Soc. It. Biol.
Sper. 55 (1989) 959^965.
[35] J.E. Pallanca, N. Smirno, J. Exp. Bot. 51 (2000) 669^674.
[36] C.J. Schoeld, Z. Zhang, Curr. Opin. Struct. Biol. 9 (1999) 722^731.
[37] J.G. Dong, J.C. Fernandez-Maculet, S.F. Yang, Proc. Natl. Acad.
Sci. USA 89 (1992) 9789^9793.
[38] K.I. Kivirikko, T. Pihlajaniemi, Adv. Enzymol. Relat. Areas Mol.
Biol. 72 (1998) 325^398.
[39] K.L. Manchester, Trends Pharmacol. Sci. 19 (1998) 167^170.
[40] J.J. Burns, Am. J. Med. 26 (1959) 740^748.
[41] N. Stone, A. Meister, Nature 194 (1962) 555^557.
[42] R.E. Rhoads, S. Udenfriend, Arch. Biochem. Biophys. 139 (1970)
329^339.
[43] L. De Gara, F. Tommasi, R. Liso, O. Arrigoni, Phytochemistry 30
(1991) 1397^1399.
[44] M.C. De Tullio, C. Paciolla, F. Dalla Vecchia, N. Rascio, S. D'Emerico, L. De Gara, R. Liso, O. Arrigoni, Planta 209 (1999) 424^434.
[45] P. Wojtaszek, C.G. Smith, G.P. Bolwell, Int. J. Biochem. Cell. Biol.
31 (1999) 463^477.
[46] J. Sommer-Knudsen, A. Bacic, A.E. Clarke, Phytochemistry 47
(1998) 483^497.
[47] B. Peterkofsky, Am. J. Clin. Nutr. 54 (1991) 1135S^1140S.
[48] D.J. Prockop, K.I. Kivirikko, N. Engl. J. Med. 311 (1984) 376^386.
[49] J.B. Cooper, J.B. Heuser, J.E. Varner, Plant Physiol. 104 (1994) 747^
752.

Cyaan Magenta Geel Zwart

O. Arrigoni, M.C. De Tullio / Biochimica et Biophysica Acta 1569 (2002) 1^9

[50]
[51]
[52]
[53]
[54]
[55]
[56]
[57]

[58]
[59]
[60]
[61]
[62]
[63]
[64]
[65]
[66]
[67]
[68]
[69]
[70]
[71]
[72]
[73]
[74]
[75]
[76]
[77]
[78]

[79]
[80]
[81]
[82]
[83]
[84]
[85]
[86]
[87]
[88]
[89]

J. Baum, B. Brodsky, Curr. Opin. Struct. Biol. 9 (1999) 122^128.

N. Hosokawa, K. Nagata, FEBS Lett. 466 (2000) 19^25.
C.J. Rebouche, Am. J. Clin. Nutr. 54 (1991) 1147S^1152S.
M. Otsuka, M. Matsuzawa, T.Y. Ha, N. Arakawa, J. Nutr. Sci.
Vitaminol. (Tokyo) 45 (1999) 163^171.
R.R. Sealock, R.L. Goodland, Science 114 (1951) 645^647.
S.C. Kumer, K.E. Vrana, J. Neurochem. 67 (1996) 443^462.
I. Garcia, M. Rodgers, R. Pepin, T. Hsieh, M. Matringe, Plant Physiol. 119 (1999) 1507^1516.
S. Jia, K. McGinnis, W.J. Van Dusen, C.J. Burke, A. Kuo, P.R.
Grin, M.K. Sardana, K.O. Elliston, A.M. Stern, P.M. Friedman,
Proc. Natl. Acad. Sci. USA 91 (1994) 7227^7231.
O. Arrigoni, M.C. De Tullio, J. Plant Physiol. 157 (2000) 481^488.
X. Qin, J.A. Zeevaart, Proc. Natl. Acad. Sci. USA 96 (1999) 15354^
15361.
J. von Lintig, K. Vogt, J. Biol. Chem. 275 (2000) 11915^11920.
M.H. Park, Y.B. Lee, Y.A. Joe, Biol. Signals 6 (1997) 115^123.
H.M. Hanauske-Abel, M.H. Park, A.R. Hanauske, A.M. Popowicz,
M. Lalande, J.E. Folk, Biochim. Biophys. Acta 1221 (1994) 115^124.
F.J. Alcain, M.I. Buron, J.C. Rodriguez-Aguilera, J.M. Villalba, P.
Navas, Cancer Res. 50 (1990) 5887^5891.
A. Hidalgo, G. Garcia-Herdugo, J.A. Gonzalez-Reyes, D.J. Morre,
P. Navas, Bot. Gaz. 152 (1991) 282^288.
P. Navas, J.M. Villalba, F. Cordoba, Biochim. Biophys. Acta 1197
(1994) 1^13.
M. Kern, E. Racker, Arch. Biochem. Biophys. 48 (1954) 235^236.
H. Beevers, Plant Physiol. 29 (1954) 265^269.
A. Nason, W.D. Wosilait, A.J. Terrell, Arch. Biochem. Biophys. 48
(1954) 233^235.
O. Arrigoni, G. Rossi, E. Marre, Rend. Accad. Naz. Lincei. Cl. Sc.
Fis. Mat. Nat. 23 (1957) 287^295.
E. Marre, O. Arrigoni, Biochim. Biophys. Acta 30 (1958) 453^457.
R.C. Green, P.J. O'Brien, Biochim. Biophys. Acta 293 (1973) 334^
342.
J.M. Villalba, F. Navarro, F. Cordoba, A. Serrano, A. Arroyo, F.L.
Crane, P. Navas, Proc. Natl. Acad. Sci. USA 92 (1995) 4887^4891.
C. Gomez-Diaz, J.M. Villalba, R. Perez-Vicente, F.L. Crane, P.
Navas, Biochem. Biophys. Res. Commun. 234 (1997) 78^81.
S. De Leonardis, G. De Lorenzo, G. Borraccino, S. Dipierro, Plant
Physiol. 109 (1993) 847^851.
C.H. Foyer, B. Halliwell, Planta 133 (1976) 21^25.
J. Bianchi, R.C. Rose, Toxicology 40 (1987) 75^82.
S. Morell, H. Follmann, M. De Tullio, I. Haberlein, FEBS Lett. 414
(1997) 567^570.
M. Fiorani, R. De Sanctis, F. Scarlatti, L. Vallorani, R. De Bellis, G.
Serani, M. Bianchi, V. Stocchi, Mol. Cell. Biochem. 209 (2000) 145^
153.
R.L. Stahl, L.F. Liebes, R. Silber, Biochim. Biophys. Acta 839 (1985)
119^121.
R.C. Rose, A.M. Bode, Enzyme 46 (1992) 196^203.
W.W. Wells, J. Bioenerg. Biomembr. 26 (1994) 420^425.
W.W. Wells, D.P. Xu, Y. Yang, J. Biol. Chem. 265 (1990) 15361^
15364.
S. Trumper, H. Follmann, I. Haberlein, FEBS Lett. 352 (1994) 159^
162.
J.B. Park, M. Levine, Biochem. J. 315 (1996) 931^938.
B. Del Bello, E. Maellaro, L. Sugherini, A. Santucci, M. Comporti,
A.F. Casini, Biochem. J. 304 (1994) 385^390.
M.P. Washburn, W.W. Wells, Biochem. Biophys. Res. Commun. 257
(1999) 567^571.
W.C. Hou, H.J. Chen, Y.H. Lin, Plant Sci. 149 (1999) 151^156.
J.G. Vethanayagam, E.H. Green, R.C. Rose, A.M. Bode, Free Radic.
Biol. Med. 26 (1999) 1591^1598.
T. Ishikawa, A.F. Casini, M. Nishikimi, J. Biol. Chem. 273 (1998)
28708^28712.

BBAGEN 25270 23-1-02

[90] Y. Kato, J. Urano, Y. Maki, T. Ushimaru, Plant Cell Physiol. 38

(1997) 173^178.
[91] T. Shimaoka, A. Yokota, C. Miyake, Plant Cell Physiol. 41 (2000)
1110^1118.
[92] M.C. De Tullio, L. De Gara, C. Paciolla, O. Arrigoni, Plant Physiol.
Biochem. 36 (1998) 433^440.
[93] A. Polle, Plant Physiol. 126 (2001) 445^462.
[94] C. Paciolla, M.C. De Tullio, A. Chiappetta, M.B. Bitonti, A.M. Innocenti, R. Liso, O. Arrigoni, Plant Cell Physiol. 48 (2001) 857^863.
[95] M. Otsuka, T. Kurata, N. Arakawa, J. Nutr. Sci. Vitaminol. (Tokyo)
32 (1986) 183^190.
[96] G. Nardai, L. Braun, M. Csala, V. Mile, P. Csermely, A. Benedetti,
J. Mandl, G. Banhegyi, J. Biol. Chem. 276 (2001) 8825^8828.
[97] J.W. Patterson, J. Biol. Chem. 183 (1950) 81^88.
[98] A.M. Bode, C.R. Yavarow, D.A. Fry, T. Vargas, Biochem. Biophys.
Res. Commun. 191 (1993) 1347^1353.
[99] M. Kashiba, J. Oka, R. Ichikawa, A. Kageyama, T. Inayama, H.
Kageyama, T. Ishikawa, M. Nishikimi, M. Inoue, S. Inoue, Biochem.
J. 351 (2000) 313^318.
[100] J.C. Deutsch, J. Chromatogr. A 881 (2000) 299^307.
[101] I. Koshiishi, Y. Mamura, T. Imanari, Biochim. Biophys. Acta 1379
(1998) 257.
[102] G. Banhegyi, L. Braun, M. Csala, F. Puskas, J. Mandi, Free Radic.
Biol. Med. 23 (1997) 793^803.
[103] T.A. Kostman, N.M. Tarlyn, F.A. Loewus, V.R. Franceschi, Plant
Physiol. 125 (2001) 634^640.
[104] R. Liso, G. Calabrese, M.B. Bitonti, O. Arrigoni, Exp. Cell Res. 150
(1984) 314^320.
[105] E. Kimoto, H. Tanaka, T. Ohmoto, M. Choami, Anal. Biochem.
214 (1993) 38^44.
[106] C.H. Jung, W.W. Wells, Arch. Biochem. Biophys. 355 (1998) 9^14.
[107] Y.Y. Shao, P.A. Seib, J.K. Kramer, D.A. VanGalen, J. Agric. Food
Chem. 41 (1993) 1391^1396.
[108] J.S. Lee, W.K. Huh, B.H. Lee, Y.U. Baek, C.S. Hwang, S.T. Kim,
Y.R. Kim, S.O. Kang, Biochim. Biophys. Acta 1527 (2001) 31^38.
[109] M.C. Djoman, J.F. Neault, S. Hashemi-Fesharaky, H.A. TajmirRiahi, J. Biomol. Struct. Dyn. 15 (1998) 1115^1118.
[110] J.M. Davidson, P.A. Lu Valle, O. Zoia, D. Quaglino Jr., M. Giro,
J. Biol. Chem. 272 (1997) 345^352.
[111] F. Pfeer, E. Casanueva, J. Kamar, A. Guerra, O. Perichart, F.
Vadillo-Ortega, Biol. Reprod. 59 (1998) 326^329.
[112] G. Seitz, S. Gebhardt, J.F. Beck, W. Bohm, H.N. Lode, D. Niethammer, G. Bruchelt, Neurosci. Lett. 244 (1998) 33^36.
[113] T. Mori, S. Itoh, S. Ohgiya, K. Ishizaki, T. Kamataki, Arch. Biochem. Biophys. 348 (1997) 268^277.
[114] A. Mizutani, N. Nakagawa, K. Hitomi, N. Tsukagoshi, Int. J. Biochem. Cell. Biol. 29 (1997) 575^582.
[115] D.R. Ganta, M.B. McCarthy, G.A. Gronowicz, Endocrinology 138
(1997) 3606^3612.
[116] M.V. Catani, A. Rossi, A. Costanzo, S. Sabatini, M. Levrero, G.
Melino, L. Avigliano, Biochem. J. 365 (2001) 77^85.
[117] G. Lopez-Lluch, M.V. Blazquez, R. Perez-Vicente, A. Macho, M.I.
Buron, F.J. Alcain, E. Munoz, P. Navas, Protoplasma 217 (2001)
129^136.
[118] N. Garcia-Muniz, J.A. Martinez-Izquierdo, P. Puigdomenech, Plant
Mol. Biol. 38 (1998) 623^632.
[119] C.F. Enloe Jr., Nutr. Today 14 (1979) 6^19.
[120] M. Caraglia, P. Tagliaferri, A. Budillon, A. Abbruzzese, Adv. Exp.
Med. Biol. 472 (1999) 187^198.
[121] M. Caraglia, M. Marra, G. Giuberti, A.M. D'Alessandro, A. Budillon, S. del Prete, A. Lentini, S. Beninati, A. Abbruzzese, Amino
Acids 20 (2001) 91^104.

Cyaan Magenta Geel Zwart